Classification of Healthcare Service Reviews with Sentiment Analysis to Refine User Satisfaction

Abstract

In natural language processing, sentiment analysis determines the polarity of a message based on lexical emotion. This technique is utilized intensively in service sectors to study the level of consumer satisfaction. However, the healthcare service field lacks such practice to detail responses in existing feedback systems. A proposed application which implements sentiment analysis is developed for improvement. User reviews are classified according to their word influences, namely positive, negative and neutral states. In addition, topic modelling is included to organize them in several service themes. A graphical user interface, GUI which records the analytical results is presented to users for interaction. This approach does not only benefit patients to choose their desired medical centres, but also healthcare management who wish to enhance their service quality.